Working Paper: NBER ID: w29313
Authors: Alexandra M. Tabova; Francis E. Warnock
Abstract: We build, from confidential security-level surveys, a novel dataset on the size, flows, coupon payments, and returns of the US Treasuries portfolios of foreign and US investors. The internally consistent dataset provides evidence on foreigners’ Treasuries portfolios that sharply contrasts with what is portrayed in the literature. Foreigners do not earn spectacularly low returns; they have higher average annual risk-adjusted returns than US investors (although the differences are well within the bounds of standard confidence intervals), as US investors’ strong mean returns are accompanied by higher volatility. Foreigners do not buy Treasuries when they are expensive; dollar-weighted returns that capture the timing and volume of purchases show that both private and official foreigners outperform US investors. And private foreign investors do not have inelastic demand; they are price sensitive, increasing purchases of Treasuries and the duration of their Treasury portfolios when non-US sovereign yields are low or decrease relative to Treasury yields. Our results should inform many literatures, including any that use data on international positions and flows. In fact, an important practical contribution is that the dataset shows which publicly available data on foreigners’ flows should (and should not) be used.
Keywords: No keywords provided
JEL Codes: F30; G11; G12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| foreign investors do not earn low returns (F21) | foreign investors achieve higher average annual risk-adjusted returns than US investors (F21) |
| non-US sovereign yields are low relative to Treasury yields (F34) | private foreign investors increase their purchases and the duration of their portfolios (F21) |
| differences in returns across investor types are driven by risk (G11) | private foreign investors achieve a return of 3.77% per annum compared to US private investors at 4.34% (F21) |
| changes in sovereign yields (H63) | foreign demand for US Treasuries (F49) |